In order to transmit data in, for example, HSDPA (High Speed Downlink Packet Access) while satisfying a desired error rate according to channel quality, adaptive modulation is under study in recent years, which measures transmission quality such as a CIR (Carrier to Interference Ratio), selects a communication mode (MCS: Modulation Coding Schemes) made up of a modulation scheme and error coding scheme based on the measured transmission quality and transmits data according to the modulation scheme and error coding scheme of the selected MCS.
For selection of MCS, a predetermined communication mode (MCS) selection table (hereinafter referred to as “table”) is used. The table is provided the correspondence between transmission quality such as a CIR and an error rate such as packet error rate (PER) and bit error rate (BER) for each MCS as shown in FIG. 1A, for example. That is, in FIG. 1A, in order to satisfy a desired error rate R, it is necessary to select MCS1 when a CIR is below a threshold a, select MCS2 when the CIR is equal to or higher than the threshold a and lower than a threshold b, select MCS3 when the CIR is equal to or higher than the threshold b and lower than a threshold c and select MCS4 when the CIR is equal to or higher than the threshold c. Therefore, the table in this case appears as shown in FIG. 1B, for example.
In the selection of MCS for data transmission, MCS which can satisfy a desired error rate is selected based on the measured transmission quality with reference to this table.
As described above, the table is based on the correspondence between the transmission quality and error rate for each MCS, but when, for example, an error occurs in a CIR measuring circuit and transmission quality cannot be measured accurately or when there is an influence from a propagation environment, the correspondence between the transmission quality and error rate may be different from the actual correspondence between the transmission quality and error rate. For this reason, the table becomes inaccurate, making it impossible to select an optimal MCS for data transmission.
To prevent such a situation, for example, the Unexamined Japanese Patent Publication No. 2002-64424 discloses a method of updating the table with correct data. This method measures transmission quality of received data, detects whether this transmission quality is different from desired transmission quality or not and rewrites the correspondence between the transmission quality and error rate when a difference is detected and updates the table.
However, the above described conventional method rewrites the table based on transmission quality of received data, and has a problem that when an amount of received data is small, it is not possible to measure the transmission quality of the received data accurately, failing to update the table correctly.
More specifically, when a base station apparatus is carrying out scheduling and data transmission to a plurality of communication terminal apparatuses, if data cannot be transmitted to a specific communication terminal apparatus, it is not possible to measure transmission quality of the communication terminal apparatus and not possible to update the table for selecting MCS of the data corresponding to the communication terminal apparatus. That is, when, for example, scheduling is performed and data is transmitted for four communication terminal apparatuses A to D as shown in FIG. 2, data 10 and data 20 are transmitted to the communication terminal apparatus D. But after data 10 is transmitted until data 20 is transmitted, the communication terminal apparatus D receives no data and cannot thereby measure transmission quality. Therefore, it is not possible to detect a difference between desired transmission quality and actual transmission quality during this period as with the above described conventional art.
Furthermore, as described above, the table is based on the correspondence between a CIR and error rate for each MCS, but if the period during which data is transmitted with a specific MCS is long, the MCS of received data is always the same and it is therefore impossible to update the correspondence between a CIR and error rate about other MCS.
When the table cannot be updated in this way, the correspondence between the CIR and error rate in the table may differ a great deal from the actual correspondence, which results in a problem that it is not possible to select an optimal MCS according to an actual channel condition.